Curr Osteoporos Rep (2011) 9:284–290 DOI 10.1007/s11914-011-0076-x

PEDIATRICS AND SKELETAL DEVELOPMENT (CRAIG LANGMAN AND MARIA LUISA BIANCHI, SECTION EDITORS)

Quantitative Computed Tomography and Computed Tomography in Children Babette S. Zemel

Published online: 4 October 2011 # Springer Science+Business Media, LLC 2011

Abstract Quantitative computed tomography (QCT) methodologies have been instrumental in deepening our understanding of bone acquisition and strength during childhood. Important publications in the last year have drawn attention to the functional muscle-bone unit, showing that factors such as population ancestry, bone size, and muscle composition are additional dimensions of bone strength that affect muscle-bone relationships. The role of adiposity in pediatric bone health is complex and may vary by sex, puberty stage, and degree of obesity. Several new studies have demonstrated the association of peripheral QCT (pQCT) outcomes with fracture, although pQCT outcomes are not superior to dual-energy x-ray absorptiometry measures in this regard. New highresolution pQCT studies document transient weakness in mid-puberty that coincides developmentally with the period of peak fracture incidence. These new studies will ultimately help us understand the development of sex differences in bone strength that emerge in adolescence. Keywords Quantitative computed tomography . QCT . pQCT . HR-pQCT . Children . Bone . Trabecular . Cortical . Bone mineral density

B. S. Zemel (*) Division of Gastroenterology, Hepatology and Nutrition, The Children’s Hospital of Philadelphia, 3535 Market Street, Room 1560, Philadelphia, PA 19104-4399, USA e-mail: [email protected] B. S. Zemel Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Introduction Quantitative computed tomography (QCT) techniques have become more widely used in pediatric research because of improvements in availability, cost, radiation exposure, and scan time. QCT techniques have distinct advantages over the more commonly used clinical densitometry method, dual-energy x-ray absorptiometry (DXA). QCT is able to separately assess cortical and trabecular bone, and it accurately determines volumetric bone mineral density (BMD), unlike DXA which uses a planar image to measure areal BMD [1]. Also, QCT can assess geometric properties of long bones and the newer high-resolution peripheral QCT (HR-pQCT) characterizes bone microstructure. In this summary, developments in QCT-based pediatric bone research over the past year are reviewed in the context of controversies and knowledge gaps in the field. Full-sized QCT devices can measure both axial and appendicular sites. Recently, published pediatric reference data for the spine facilitate the use of this technique [2]. Most QCT studies in children use table-top peripheral QCT (pQCT) devices to measure the radius, tibia, or femur. Measurements in long bones of growing children are complicated by the presence of growth plates, rendering identification of reference points and optimal measurement sites an arduous task. There is a lack of consistency in the literature on these particulars of pQCT methodology, and virtually no evidence on which to base protocol choice [3]. In addition, measurements on the arm versus the leg are only moderately correlated, as physical activity patterns differentially affect loading on these bone [4]. Study results must be interpreted in light of measurement site and its relevance to the hypothesis under investigation. Despite these limitations, pQCT devices provide a wealth of information about the composition and structure

Curr Osteoporos Rep (2011) 9:284–290

285

of bone. At metaphyseal sites, total and trabecular volumetric bone mineral density (vBMD) and bone area can be determined. At mid-shaft sites, cortical bone is characterized by measures such as cortical content, area, vBMD and thickness, endosteal and periosteal circumferences, and derived measures of structural strength including crosssectional moment of inertia, sectional modulus, and the strain strength index. To some degree, these measures are interrelated but each provides unique information about the development of bone strength in the growing skeleton. HR-pQCT is a new technology that is not widely available. Small regions of the tibia or radius are measured, so the effective radiation dose is low [5]. Outcomes include total bone area, total, trabecular, and cortical vBMD, cortical thickness, and bone strength index; microarchitecture parameters include cortical porosity, trabecular bone volume fraction (the ratio of bone volume to total volume), trabecular number, trabecular thickness, and trabecular separation. A comparison of imaging techniques and their strengths and limitations is shown in Table 1.

New Developments in the Functional Muscle-Bone Unit The mechanostat model posits that bones respond to the peak forces created by muscle contraction. Bones and muscles form a “functional unit” because of the close relationship between bone strength and muscle force. This conceptual model identifies “primary” bone defects, where bone strength is inadequate for muscle size; versus “secondary” or physiologic bone disorders, where bone and muscle strength are both low [6]. Strong cross-sectional associations between muscle size and bone strength are evident in healthy children, but longitudinal associations in

both healthy children [7, 8] and children with chronic medical conditions [9, 10] are not consistent with this model. Further studies in varied populations and using sophisticated statistical modeling are directing the refinement of this conceptual model of bone health and development. Several years ago, Wetzsteon et al. [11] examined ethnic differences in bone strength of the radius and tibia in healthy US children (n=63), 9 to 12 years of age, of European and African ancestry and Hispanic ethnicity. Children of European ancestry had significantly lower bone strength (polar strain strength index and bone strength index) than African American and Hispanic children after adjusting for age, sex, limb length, and muscle crosssectional area. Lower vBMD and cortical bone area contributed to the lower bone strength in the Caucasian group. Given that muscle area is a good surrogate measure of mechanical load on bone, these findings suggest that Caucasians have lower bone strength relative to the loads to which they are exposed, and thus, a lower mechanosensitivity compared with the African American and Hispanic children. A primary limitation of this analysis was the inability to adjust for potential group differences in pubertal timing. A subsequent publication by Leonard et al. [12•] assessed the independent effects of sex, population ancestry, and puberty on the functional muscle-bone unit across a broader age range (5–35 years) in 665 healthy participants (306 African Americans). Cortical bone mineral content (BMC) and vBMD, periosteal and endosteal circumferences, section modulus (38% site), and muscle area (66% site) of the tibia were assessed. Adjusting simultaneously for tibia length, age, population ancestry, sex, and Tanner stage, African Americans had significantly (P≤0.001) greater

Table 1 Comparison of imaging techniques and their strengths and limitations

Availability Cost of scans Radiation exposurea Speed Ability to measure trabecular density Ability to measure cortical density Ability to measure bone geometry Ability to measure microstructure a

DXA

QCT

pQCT

HR-pQCT

Widely available, dedicated to bone densitometry Low Low Fast No

Widely available in hospitals, not dedicated to bone densitometry Moderate Moderate Fast Yes

Not widely available, mostly research dedicated Research only Low Moderately fast Yes

Available at only a few research centers Research only Low Fast Yes

Only at cortical sites such as forearm Only with hip structural analysis No

Yes

Yes

Yes

Yes

Yes

Yes

No

No

Yes

For more information see Damilakis et al. [5]

DXA dual-energy x-ray absorptiometry; HR-pQCT high-resolution peripheral quantitative computed tomography; pQCT peripheral quantitative computed tomography; QCT quantitative computed tomography

286

values for all cortical measures in Tanner stages 1 through 4, but differences were nonsignificant in Tanner stage 5. Females had lower cortical BMC, periosteal circumference, and section modulus than males in all Tanner stages (all P<0.001). Adjustment for muscle area attenuated but did not eliminate sex and population ancestry differences in cortical dimensions. Fat area was not independently related to bone outcomes. Most importantly, they found the interaction term for muscle area x population ancestry group was nonsignificant, meaning that a unit increase in muscle area was associated with a similar increase in bone strength among African Americans and non-African Americans. In other words, adjusted for all other factors, sex and population ancestry group had independent effects that were consistent across the spectrum of muscle area. A possible explanation for these findings is that the muscle-specific forces acting on bone differ according to population ancestry and sex. Wetzsteon et al. [13•] tested this hypothesis in a subset of these participants. Peak dorsiflexion muscle torque (ft-lb) of the ankle was measured using isometric dynamometry in 321 subjects, 5 to 35 years of age, and average hours per week of physical activity was estimated by questionnaire. Muscle-specific force was determined as torque relative to muscle area.Tibia cortical bone measures similar to those described above were evaluated. Muscle-specific force was greater in the African American group (P<0.05) compared with the nonAfrican American group, even when adjusted for age and tibia length. Females had lower values than males (P < 0.001). Section modulus and other cortical bone outcomes were examined in multivariate models that demonstrated the independent and significant (all P<0.02) effects of tibia length, muscle area, muscle torque, body weight, and physical activity. Fat area was not independently related to bone outcomes. These multiple measures of bone loading attenuated but did not eliminate the significant differences between the African American and non-African American groups. These findings further support the concept that population differences in bone strength are independent of mechanical loads. Further refinement of the “mechanostat theory” is required to explain these differences. This latter study had two other important findings. First, tibia length and muscle area were similar in their strong association with section modulus, a composite measure of bone strength. A multivariate model showed that both of these dimensions were independently related to bone strength (R2 =0.91 for subjects <22 years of age) and of equal explanatory value; the partial R2 for muscle area was 0.42, adjusted for tibia length, and the partial R2 for tibia length was 0.37, adjusted for muscle area. Similar results were found in the adult age range. Thus, the functional

Curr Osteoporos Rep (2011) 9:284–290

muscle-bone unit in growing children should be revised to include bone size as an additional dimension. Second, muscle force and muscle area were comparable in explanatory value in their models. They suggested that tibia cross-sectional muscle area assessed by pQCT is an excellent indicator of muscle force in healthy subjects. However, a recent study suggests that muscle composition may be an important consideration in addition to muscle size. As adiposity increases, the amount of intramyocellular and extramyocellular fat also increases [14], so muscle composition varies across the spectrum of adiposity. In a study of 444 healthy girls, ages 9 to 12 years, Farr et al. [15••] examined muscle density using pQCT and its relationship to bone strength of the tibia and femur. Lower muscle density reflects a higher fat content in muscle. Muscle density was inversely correlated to subcutaneous adipose tissue and total body fat. Adjusting for muscle area, bone length, maturity, and ethnicity, muscle density was associated with trabecular vBMD and bone strength index at metaphyseal regions, and cortical vBMD, area, and thickness in diaphyseal regions of the tibia and femur. In the multivariate models, subcutaneous adipose tissue was not associated with bone outcomes except for bone strength index at the 4% tibia site (β =0.09, P=0.03). These important findings argue the need to consider muscle composition in the functional muscle-bone unit model.

Adiposity and Bone Strength The increased prevalence of pediatric obesity has led to concerns about its effect on bone health. Increased fracture rates have been reported for heavier children [16, 17]. However, there is conflicting evidence regarding the relationship between fat mass and bone strength in childhood. Differences in study design, analytic approaches, and sample characteristics (age and puberty stage) complicate comparison across studies. Obese children have greater fat mass accompanied by greater lean mass, yet the functional muscle-bone unit model ignores fat. Body fat and its distribution change with pubertal progression in a sexdependent fashion, and body fat may be related to the timing of puberty in girls. Consequently, the same amount of body fat in a prepubertal child may have different implications for bone strength than for a child well advanced in puberty. Visceral fat is associated with increased proinflammatory cytokines and adipokines, which can have independent effects on bone. Thus, not all fat is the same. The studies reviewed here address these complex issues. In a cross-sectional cohort (n=172) of 6-year olds in the United Kingdom, Cole et al. [18] evaluated tibia trabecular (4% site), cortical vBMD (38% site), and geometric measures, and total fat and lean mass assessed by DXA.

Curr Osteoporos Rep (2011) 9:284–290

Fat mass (adjusted for lean mass) was inversely associated with trabecular (P<0.01) and cortical vBMD (P<0.05), but positively associated with bone size (total bone area, cortical area, and cortical content) at the 38% site. Failure to account for sex and tibia length make it difficult to interpret these findings. As noted above in the studies by Leonard et al. [12•], Wetzsteon et al. [13•], and Farr et al. [15••], they found no effect of fat mass on bone outcomes adjusted for muscle area in multivariate models that adjusted for tibia length (and other factors), which may account for the conflicting results. However, this study by Cole et al. [18] focused on a young, prepubertal, and narrowly defined age range, which may in part explain their finding. The effect of fatness on bone outcomes may be nonlinear. In a study of 186 healthy children, 7 to 19 years of age in Finland, Viljakainen et al. [19] divided the group into tertiles based on age and sex-specific values for percent body fat by DXA. Multivariate models adjusting for lean mass, pubertal development, dietary calcium intake, and physical activity showed a signficant difference in cortical vBMD (66% radius) such that the intermediate fatness group had the highest values (P=0.02). However, these findings were not adjusted for the known age and sexrelated changes in cortical vBMD in this age range making it difficult to draw firm conclusions from these findings. Farr et al. [20] examined the relationship between total fat mass by DXA and pQCT measures of the femur (4% and 20% sites) and tibia (4% and 66% sites) in their crosssectional cohort of US girls (n=396; 8–13 years of age). Body fat was positively associated with bone outcomes in unadjusted analyses. Multivariate models adjusting for muscle area, puberty, bone length, physical activity, and ethnicity found no significant association between fat mass and any bone parameters. In a final analysis, they adjusted for the potential force of impact during a fall by dividing all bone outcomes by body mass; they compared groups based on tertiles of total body fat and adjusted for the same covariates as above. All bone outcomes expressed as a ratio with body mass were significantly (P<0.001) lower in the tertile with greatest fat mass compared with the middle and lowest tertiles. However, this final analysis should be interpreted with extreme caution for two reasons. First, the bone outcomes were divided by body weight, yet the independent variables included both fat and muscle, which largely comprise body weight; they effectively added body weight (fat and muscle) to one side of their equation and the inverse of body weight to the other, which may explain the inverse trend. Second, the use of ratios (eg, bone outcome divided by body mass) in biological outcomes should always be used with caution because it assumes a specific linear relationship between two variables such that there is a zero intercept and a slope of one [21].

287

Nutrition and Bone Few studies have examined the effects of nutrition on bone strength in children. With the growing concern about vitamin D sufficiency, the effects of vitamin D status on bone vBMD and strength have been a topic of investigation. In one of the first studies to use pQCT in infants, Viljakainen et al. [22] assessed vitamin D status during pregnancy and acquired pQCT scans of the distal tibia (20% site) close to birth and at 14 months of age in 86 infants. Maternal vitamin D status during pregnancy was used to divide infants into low versus high vitamin D groups. At birth, the infants’ 25-hydroxyvitamin D levels were 36.3 versus 52.5 nmol/L (P<0.001) in the low versus high groups, but there was no difference at 14 months (63 vs 66 nmol/L; P=0.58). At birth, infants in the low group had significantly lower tibia BMC and cross-sectional bone area than the high group, but gains in BMC over 14 months resulted in similar BMC at 14 months. The greater bone area in the high group at birth persisted at 14 months (multivariate analysis of variance; P=0.068). There was no difference in vBMD or change in BMD between the two groups. These finding suggest that prenatal vitamin D exposure may have long-lasting effects on the growing bone, despite normalization of vitamin D status. Ward et al. [23] conducted a double-blind, randomized controlled vitamin D supplementation trial in 12- to 14-year-old postmenarchal girls (n=73) in the United Kingdom. Participants were given four doses of 150,000 IU of ergocalciferol over 1 year. The treatment resulted in significant improvement in 25-hydroxyvitamin D status (56.0±8.9 nmol/L vs 15.8±6.6 nmol/L), but there were no effects of supplementation on any bone outcomes assessed by pQCT at radius (4% and 50% sites) or tibia (10% and 66% sites). Further study is needed to determine whether factors such as dose regimen, calcium intake, skeletal maturation, and physical activity influence these findings. A randomized controlled vitamin D plus calcium supplementation trial in peripubertal female identical twins (n=20; ages 9 to 13 years) showed a more positive impact on bone outcomes [24]. One of each pair was randomized to receive 800 mg of calcium and 400 IU of cholecalciferol, or placebo for 6 months. pQCT measurements were obtained at the tibia (4%, 14%, 38%, and 66% sites) and radius (4% and 66% sites). The supplement group had significantly greater (P=0.001) increases in tibia and radius trabecular vBMD, area, and strength strain index and greater (P=0.001) increases in cortical area at the 38% and 66% of tibial sites over 6 months. Both trabecular and cortical bone responded to calcium and vitamin D supplementation in this age range. These three studies suggest that

288

during periods of rapid growth, the immature skeleton may be more sensitive to the effects of vitamin D status on bone health than during periods of slow growth.

Fracture A decade ago, Skaggs et al. [25] used QCT to examine differences in the forearm of 50 girls, ages 4 to 15 years, who had sustained a forearm fracture compared with those who had not, matched for age, height, weight, and Tanner stage. They found smaller cross-sectional area of the distal radius (8%) in the fracture group compared with controls, but no differences in cortical or trabecular density. This study was the first to demonstrate that “size matters” with respect to bone strength during childhood. In a retrospective fracture study of 465 girls, ages 8 to 13 years, Farr et al. [26] assessed bone strength of the tibia (4% and 66% sites) and femur (4% and 20% sites) by pQCT and body composition by DXA. A total of 19% had at least one fracture, and 3% had more than one fracture. The majority (49%) were forearm fractures; all fractures were included in the analysis. Controlling for maturity, body mass, leg length, ethnicity, and physical activity, lower trabecular vBMD at distal metaphyseal sites of the femur and tibia was significantly associated with prior fracture; each standard deviation decrease in trabecular vBMD at the distal femur and tibia was associated with a 1.4 (1.1–1.9) and 1.3 (1.0–1.7) increased risk of fracture. No other pQCT outcomes, including cross-sectional area, were associated with fracture. The DONALD (Dortmund Nutritional and Anthropometric Longitudinally Designed) study of healthy children (ages 5–17 years) in Germany examined the relationship between pQCT measurements of cortical and trabecular bone of the radius to fractures reported over the following 5 years (n=229). A total of 35% of the sample reported at least one fracture in the 5-year period; the majority were forearm fractures. There were no differences in age, body size (height, weight, body mass index), or bone outcomes (cortical and trabecular vBMD, muscle area, or strain strength index) between the fracture and nonfracture groups. The large age range and varied fracture sites may be responsible for the lack of association because only a subset of subjects would be of an age when bone outcomes might be associated with fracture. Bone outcomes may be more strongly associated with fracture in the peripubertal age range when fractures are most common. Cheng et al. [27] enrolled girls (n=396) at 10 to 13 years of age and followed them longitudinally for up to 8 years. vBMD and geometry of the distal radius were assessed. Lifetime fracture history, verified by medical records, was obtained from birth to 20 years. Fracture

Curr Osteoporos Rep (2011) 9:284–290

incidence peaked at ages 8 and14 years; 38% were upper limb fractures. Upper limb fracture in this age range was associated with lower distal radial vBMD (4% site) at baseline and up to 7 years later. These findings suggest that periods of bone fragility exist and these fractures in particular are most likely to be characterized by lower vBMD that persists to young adulthood. A retrospective study of young men (n=1,068) in Sweden showed similar results, with lower trabecular vBMD of the radius (6.6%) and tibia (4.5%) and slightly lower cortical vBMD of the radius (0.4%) and tibia (0.3%) in those with a history of fracture compared with men who did not fracture [28]. For both the radius and the tibia, each standard deviation decrease in trabecular vBMD was associated with a 1.46 increased fracture prevalence. A major obstacle in group comparisons is similarity of exposure. In an elegant case-controlled study, Kalkwarf et al. [29••] compared healthy children (5–16 years of age) who experienced a forearm injury resulting in a fracture (n=224), to children who experienced a forearm injury without fracture (n=200). pQCT forearm measurements were obtained and expressed as Z-scores adjusted for age, sex, race, height, and Tanner stage. Compared with controls, the fracture group had significantly (P<0.05) lower total vBMD (−3.2%) at the distal radius and lower total BMC (−3.4%), cortical vBMD (−0.9%), cortical area (−2.8%), and strain strength index (−4.6%) at the mid-shaft. Results for DXA outcomes were in a similar range, and the odds ratio associated with a one standard deviation decrease was 1.28 (95% CI, 1.03, 1.59) to 1.41 (95% CI, 1.07, 1.85) for all bone outcomes. These odds ratios are remarkably similar to those described above as reported by Farr et al. [26] and Cheng et al. [27].

HR-Pqct HR-pQCT (XtremeCT; Scanco Medical; Brüttisellen, Switzerland) is a new technology that will ultimately reveal the microarchitectural changes in bone during growth, but at present, findings are limited. HR-pQCT scans are rapid (3 min) with a relatively low radiation dose (<3 μSv) [30]. The first report in children was a cross-sectional study of the radius in 127 children, ages 6 to 21 years [31]. Subjects were grouped according to bone age into five puberty groups. Images were acquired starting at 1 mm distance from the proximal limit of the epiphyseal growth plate and spanned 9.02 mm. The findings showed greater trabecular number, thickness, and bone volume fraction among boys more advanced in puberty compared with those less advanced, but no such differences were observed in girls. Cortical thickness and density were modestly lower (girls) or similar (boys) in mid-puberty compared with prepuberty;

Curr Osteoporos Rep (2011) 9:284–290

higher levels occurred at the end of puberty in both sexes. Total bone strength increased steadily across puberty groups. However, the ratio of cortical to trabecular bone volume was lower in mid- compared with late puberty groups in both sexes, and cortical porosity peaked during this time. This possible decline in bone strength of the radius may correspond to a transient weakness in bone that results in the increase in forearm fracture rate at these developmental stages. Wang et al. [32] observed similar patterns at the distal radius (4% site) and tibia (7% site). Although crosssectional, these data suggest that the transient weakness in cortical vBMD at the distal radius in mid-puberty was not evident in the tibia. They found similar values for trabecular number across puberty stages for both sexes and measurement sites, but gains in trabecular thickness with advancing puberty stage in males resulted in an overall increase in bone volume fraction in both sexes, but more so in males. Burrows et al. [33] assessed tibia microstructure (8% site) in 15- to 20-year-old healthy subjects (n=279). Total and cortical vBMD, cortical thickness, and bone strength index increased across the age range in both males and females, but no age-related changes in total area, trabecular density, or bone volume fraction. Males had significantly greater values for most outcomes, although females had significantly greater cortical vBMD. Liu et al. [4] examined the radius (7% site) and tibia (8% site) (n=323; 9–21 years of age) and demonstrated discordance in bone architecture parameters (Z-scores) between sites. Heterogeneity in Z-scores was related to body size variables, implying that weight-bearing has a greater influence on tibia outcomes compared with the radius. Among 15 to 20-year olds, McKay et al. [34] showed that load-bearing physical activity was associated with moments of inertia in males, and trabecular and total vBMD in females in the distal tibia (8% site). Combined, these studies add to the mounting evidence of transient cortical weakness in mid-puberty, particularly for the radius, and site-specific gender- and puberty-related differences in bone architecture and strength. However, longitudinal studies are needed to confirm that these crosssectional patterns occur within individuals and identify the causes and consequences of skeletal fragility during growth.

Conclusions QCT methodologies have been instrumental in deepening our understanding of bone acquisition and strength during childhood. Important publications in the last year have drawn attention to important considerations in the use of the functional muscle-bone unit to categorize bone disease as a

289

primary or secondary disorder. Factors such as population ancestry, bone size, and muscle composition are additional dimensions of bone strength that affect muscle-bone relationships. The controversy continues in identifying the role of adiposity in pediatric bone health; does fat mass have an additional impact on bone strength after accounting for bone size and muscle, and does it vary by sex, puberty stage, and degree of obestiy? Several new studies have demonstrated the association of pQCT outcomes with fracture, although pQCT outcomes are not superior to DXA measures in this regard. Vitamin D status is important for bone development in the newborn and peripubertal periods, but may have little impact on postmenarchal females. New high-resolution pQCT studies are emerging that document transient weakness in mid-puberty, which coincides developmentally with the period of peak fracture incidence. These new studies will ultimately help us understand the development of sex differences in bone strength that emerge in adolescence.

Disclosure Conflicts of interest: B. Zemel: has received a consulting fee or honorarium from the International Society for Clinical Densitometry for a lecture in the ISCD Pediatric Bone Densitometry course on pQCT in children; and has received grant support from the National Institutes of Health for studies in which she was a co-author.

References Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance 1. Leonard MB, Zemel BS. Current concepts in pediatric bone disease. Pediatr Clin North Am. 2002;49:143–73. 2. Gilsanz V, Perez FJ, Campbell PP, et al. Quantitative CT reference values for vertebral trabecular bone density in children and young adults. Radiology. 2009;250:222–7. 3. Zemel B, Bass S, Binkley T, et al. Peripheral quantitative computed tomography in children and adolescents: the 2007 ISCD Pediatric Official Positions. J Clin Densitom. 2008;11:59–74. 4. Liu D, Burrows M, Egeli D, McKay H. Site specificity of bone architecture between the distal radius and distal tibia in children and adolescents: An HR-pQCT study. Calcif Tissue Int. 2010;87:314–23. 5. Damilakis J, Adams JE, Guglielmi G, Link TM. Radiation exposure in X-ray-based imaging techniques used in osteoporosis. Eur Radiol. 2010;20:2707–14. 6. Schoenau E. From mechanostat theory to development of the “Functional Muscle-Bone-Unit”. J Musculoskelet Neuronal Interact. 2005;5:232–8. 7. Wey HE, Binkley TL, Beare TM, et al. Cross-sectional versus longitudinal associations of lean and fat mass with pQCT bone outcomes in children. J Clin Endocrinol Metab. 2011;96:106–14.

290 8. Xu L, Nicholson P, Wang Q, et al. Bone and muscle development during puberty in girls: a seven-year longitudinal study. J Bone Miner Res. 2009;24:1693–8. 9. Burnham JM, Shults J, Sembhi H, et al. The dysfunctional muscle-bone unit in juvenile idiopathic arthritis. J Musculoskelet Neuronal Interact. 2006;6:351–2. 10. Dubner SE, Shults J, Baldassano RN, et al. Longitudinal assessment of bone density and structure in an incident cohort of children with Crohn’s disease. Gastroenterology. 2009;136:123–30. 11. Wetzsteon RJ, Hughes JM, Kaufman BC, et al. Ethnic differences in bone geometry and strength are apparent in childhood. Bone. 2009;44:970–5. 12. • Leonard MB, Elmi A, Mostoufi-Moab S, et al. 2010 Effects of sex, race, and puberty on cortical bone and the functional muscle bone unit in children, adolescents, and young adults. J Clin Endocrinol Metab 95:1681–1689. This cross-sectional study examined age-related patterns in tibia cortical bone outcomes (BMC, vBMD, periosteal and endosteal circumferences, and section modulus) and muscle-bone relationships. Greater bone strength in males versus females, and blacks versus nonblacks was evident after adjusting for muscle area, but patterns varied by puberty stage. 13. • Wetzsteon RJ, Zemel BS, Shults J, et al. 2011 Mechanical loads and cortical bone geometry in healthy children and young adults. Bone 48:1103–1108. This study examined the effects of mechanical load and muscle area on tibia cortical bone geometry in healthy subjects ages 5 to 35 years. Muscle cross-sectional area, muscle strength, body weight, and physical activity were independently and significantly associated with cortical bone outcomes, adjusting for age, tibia length, puberty stage, sex, and African ancestry. Differences between black and nonblack groups in cortical bone strength were independent of differences in maturational timing, muscle area, and strength. 14. Sinha R, Dufour S, Petersen KF, et al. Assessment of skeletal muscle triglyceride content by (1)H nuclear magnetic resonance spectroscopy in lean and obese adolescents: relationships to insulin sensitivity, total body fat, and central adiposity. Diabetes. 2002;51:1022–7. 15. •• Farr JN, Funk JL, Chen Z, et al. 2011 Skeletal muscle fat content is inversely associated with bone strength in young girls. J Bone Miner Res pQCT was used to assess subcutaneous adipose tissue (SAT) and muscle density, an index of skeletal muscle fat content of the femur and tibia. SAT was positively correlated and muscle density was negatively correlated with total body fat mass. Muscle density was associated with indices of bone strength at metaphyseal and diaphyseal regions of the femur and tibia. 16. Goulding A, Jones IE, Taylor RW, et al. Bone mineral density and body composition in boys with distal forearm fractures: a dualenergy x-ray absorptiometry study. J Pediatr. 2001;139:509–15. 17. Clark EM, Ness AR, Tobias JH. Bone fragility contributes to the risk of fracture in children, even after moderate and severe trauma. J Bone Miner Res. 2008;23:173–9. 18. Cole ZA, Harvey NC, Kim M, et al. 2011 Increased fat mass is associated with increased bone size but reduced volumetric density in pre pubertal children. Bone 19. Viljakainen HT, Pekkinen M, Saarnio E, et al. Dual effect of adipose tissue on bone health during growth. Bone. 2011;48:212–7.

Curr Osteoporos Rep (2011) 9:284–290 20. Farr JN, Chen Z, Lisse JR, et al. Relationship of total body fat mass to weight-bearing bone volumetric density, geometry, and strength in young girls. Bone. 2010;46:977–84. 21. Allison DB, Paultre F, Goran MI, et al. Statistical considerations regarding the use of ratios to adjust data. Int J Obes Relat Metab Disord. 1995;19:644–52. 22. Viljakainen HT, Korhonen T, Hytinantti T, et al. Maternal vitamin D status affects bone growth in early childhood–a prospective cohort study. Osteoporos Int. 2011;22:883–91. 23. Ward KA, Das G, Roberts SA, et al. A randomized, controlled trial of vitamin D supplementation upon musculoskeletal health in postmenarchal females. J Clin Endocrinol Metab. 2010;95:4643– 51. 24. Greene DA, Naughton GA. Calcium and vitamin-D supplementation on bone structural properties in peripubertal female identical twins: a randomised controlled trial. Osteoporos Int. 2011;22:489– 98. 25. Skaggs DL, Loro ML, Pitukcheewanont P, et al. Increased body weight and decreased radial cross-sectional dimensions in girls with forearm fractures. J Bone Miner Res. 2001;16:1337–42. 26. Farr JN, Tomas R, Chen Z, et al. Lower trabecular volumetric BMD at metaphyseal regions of weight-bearing bones is associated with prior fracture in young girls. J Bone Miner Res. 2011;26:380–7. 27. Cheng S, Xu L, Nicholson PH, et al. Low volumetric BMD is linked to upper-limb fracture in pubertal girls and persists into adulthood: a seven-year cohort study. Bone. 2009;45:480–6. 28. Darelid A, Ohlsson C, Rudang R, et al. Trabecular volumetric bone mineral density is associated with previous fracture during childhood and adolescence in males: the GOOD study. J Bone Miner Res. 2010;25:537–44. 29. •• Kalkwarf HJ, Laor T, Bean JA 2011 Fracture risk in children with a forearm injury is associated with volumetric bone density and cortical area (by peripheral QCT) and areal bone density (by DXA). Osteoporos Int 22:607–616. This study compared DXA and pQCT outcomes in children 5 to 16 years of age with forearm fractures to children injured without fracture. Fracture cases had lower total vBMD but not trabecular vBMD or bone area at the 4% radius site, and lower cortical vBMD, cortical area, and strength strain index at the 20% radius site. Similar patterns were observed for DXA outcomes of the radius, spine, and hip. 30. Burrows M, Liu D, Perdios A, et al. Assessing bone microstructure at the distal radius in children and adolescents using HRpQCT: a methodological pilot study. J Clin Densitom. 2010;13:451–5. 31. Kirmani S, Christen D, van Lenthe GH, et al. Bone structure at the distal radius during adolescent growth. J Bone Miner Res. 2009;24:1033–42. 32. Wang Q, Wang XF, Iuliano-Burns S, et al. Rapid growth produces transient cortical weakness: a risk factor for metaphyseal fractures during puberty. J Bone Miner Res. 2010;25:1521–6. 33. Burrows M, Liu D, Moore S, McKay H. Bone microstructure at the distal tibia provides a strength advantage to males in late puberty: an HR-pQCT study. J Bone Miner Res. 2010;25:1423– 32. 34. McKay H, Liu D, Egeli D, et al. Physical activity positively predicts bone architecture and bone strength in adolescent males and females. Acta Paediatr. 2011;100:97–101.